This PDF is a selection from an out-of-print volume from the National
Bureau of Economic Research
Volume Title: The Growth of American Trade Unions, 1880-1923
Volume Author/Editor: Leo Wolman
Volume Publisher: NBER
Volume ISBN: 0-87014-005-1
Volume URL: http://www.nber.org/books/wolm24-1
Publication Date: 1924
Chapter Title: Chapter IV. The Extent of Labor Organization in 1910
and 1920
Chapter Author: Leo Wolman
Chapter URL: http://www.nber.org/chapters/c5903
Chapter pages in book: (p. 82 - 96)
CHAPTER IV
THE EXTENT OF LABOR ORGANIZATION IN
1910 AND
Even the number of wage earners, as defined in the last chapter,
would not be considered by some a thoroughly fair base for measuring the achievement in size of• an organized labor movement.
The final figure there derived includes agricultural employees, whom
trade unions have made little effort to organize, and such other
groups, like clerical workers, whose adherence to the trade union
is of comparatively recent origin. Trade unions, moreover, limit
their membership in a great variety of ways. Almost all exclude
persons not yet of a specified age; some have standards of skill
which prospective members must meet; others impose high initiation fees or require attachment to the industry for a specified period
of years; and still others impose restrictions on entry based on the
color and sex of the applicants to membership. The extent of
trade unionism would naturally appear in its most favorable statistical light, if allowance were made for these various factors and
trade union membership were compared only with the residuum of
organizable employees. No attempt is, however, here made to deal
with such refinements; and comparison is always made between
the numbers in trade unions and the numbers of those employees,
who are, by common consent, regarded as likely material for organization in trade unions.
Computing the percentage that union membership represents of
the total number of wage earners in the United States and of the
number in the major and minor divisions of industry involves technical difficulties, which cannot be altogether overcome. The most
serious of these is encountered in the attempt to make the statistics
of membership conform to the statistics of occupations. Union
membership, since many unions are organized along trade or occupational lines, frequently overlaps the industrial classes of the census.
82
THE EXTENT OF LABOR ORGANIZATION
83
Although nearly all labor organizations have most elaborate official statements of their jurisdictional claims, it is generally impossible, except by arbitrary decision, to split their membership among
the various industrial classifications of the census. The carpenters'
union, for instance, as has already been pointed out, includes in
its membership not only building carpenters but also factory workers
employed in the industry, described by the census as CC lumber
and its manufactures." 'While it is possible to effect a distribution, in round numbers, of the total membership of this union into
the number working on buildings and those in factories, finer estimates are impracticable. The same holds true of other important
The difficulty encountered in separating the self-employed
from those employed by others, which is a real one in many industries and occupations, has already been discussed.
unions.
The detailed materials for computing the percentage organized
among the various divisions of industry and among selected occupations and the results of those computations are presented in the
Appendix to this volume in Tables V, VI, VII, VIII and IX. The
first table of this series shows the membership of every American
national and international union in the year 1920. It differs from
the first exhaustive table of membership (Table I) in that it contains also the Canadian membership of each American organization
which has jurisdiction over workers in Canadian industry. From
it are derived the statistics of the membership in the United States
alone of the unions included in the table. Since the occupation
census gives statistics only for the continental United States, comparison can properly be made only with the United States membership of labor organizations.
Actual comparisons between membership and the number of
wage earners in the year 1920 are shown in Table VI. This table
was constructed by separating for each industry and subdivision
of industry in the census of occupations of 1920 the wage earners
from the salaried and employer classes. Next to these figures were
placed the statistics of trade union membership in the United
States taken from Table V. It will be found that the number of
wage earners in certain industries differs substantially from the statistics of Table IV. This is due largely to the fact that the juris-
diction claims of the unions forced the inclusion in Table VI of
THE GROWTH OF TRADE UNIONS, 1880-4923
84
several categories of workers who were clearly employed in managerial or supervisory functions and who were, consequently, properly classified in Table IV in the salaried group. The final percentages of organization underestimate somewhat the prevailing
extent of organization in 1920 because of the omission of independent local unions, whose membership could not be obtained,
and of local unions directly affiliated with the American Federation
of Labor, whose membership could not, through the lack of the
necessary data, be distributed among the industries of the country.
The net effects of these omissions are almost imperceptible and are
not likely to change the present results by any more than one
per cent.1
Similar statistics showing the percentage of organization in 1910
are given in Table VII. This table is reprinted, substantially in
its original form, from the article already cited.2 It has, however,
been revised in several important particulars to meet differences
between the census of 1910 and that of 1920. It has been pointed
out before that the occupation statistics of 1910 were in much more
detailed form than those of 1920. It was possible in the earlier
census to make finer classifications than could be used in dealing
with the later census. For this reason consolidated classifications
replaced the more detailed ones in the original table. In some
instances apparent improvements in classification dictated modifications in 'the original 1910 table. Thus coopers were taken from
the "hand trades," which are altogether omitted from the revised
table, and were put into the lumber and furniture industries. The
large category of electric light and power plants, electric supply
factories, electricians and electrical engineers, telegraph and teleThe membership of the I. W. W. was, also, omitted from the table because it was
made available too late. The office of the .1. W. W. reports its membership to have
been distributed in 1920 as follows:
7,000
Lumber Workers
6,000
Agricultural Workers
4,600
Mine Workers
5,500
General Construction Workers....
4,700
Railroad Workers
3,800
House and Building Construction Workers
4,000
Metal and Machinery Workers. ..
6,000
Marine Transport Workers
1
Total
Quarterly Journal of Economics, May, 1916, p. 606.
41,600
THE EXTENT OF LABOR ORGANIZATION
85
phone linemen, was split into a number of diverse elements and the
membership of the electrical workers' union, originally allotted to
this comprehensive group, was reapportioned to the new industrial classes. The most radical revisions in the original table were
made for the general occupational divisions of "trade," "public
service," "professional service," "domestic and personal service,"
"agriculture, forestry and animal husbandry," "proprietary, supervisory and official" groups, and "clerical" groups. All of these
groupings and classifications contained in the original table were
completely discarded and were replaced by revised figures for both
1910 and 1920, taken entirely from the census of occupations of
1920. In their present form, Tables VI and VII, showing respectively percentages of labor organization in 1920 and 1910, possess
a high degree of comparability.
Compared with the total numbers of wage earners in this country, trade union strength as measured by its membership was relatively twice as great in 1920 as in 1910. The rate of growth during this decade was approximately the same whether membership
is compared, with the industrial wage-earning population of the
country or with the combined industrial and agricultural wageearning population. The tabulation below shows that, roughly, oneTOTAL
WAGE
EARNERS
TRADE
UNION
MEMBERSHIP
PER CENT
ORGANIZED
IN U. S.
1920...
1910...
26,080,689
22,406,714
4,881,200
2,101,502
18.7
9.4
TOTAL
WAGE
EARNERS
EXCLUDING
AGRICULTURE
TRADE
UNION
MEMBERSHIP
23,480,077
19,262,941
4,881,200
2,101,502
PER CENI
ORGANIZED
IN U. S.
20.8
10.9
fifth of the wage earners of the country were members of labor
organizations; whereas in 1910 something like one-tenth were so
In other words in 1920, after 10 years of very substantial growth in numbers, about four-fifths of the general category
of wage earners were not members of unions. While these figures
are in themselves of considerable interest, their full significance
cannot be clear without detailed inquiry into the sources of union
strength and weakness. Such inquiry can be conducted only by
organized.
86,
THE GROWTH OF TRADE UNIONS, 1880-1923
discovering the varying magnitude of trade unionism in the great
divisions of industry.
It has long been generally appreciated that labor organizations
receive their first impetus and make their most striking headway
among the so-called manual workers, those who work in factories
and mines, on railroads and buildings; and that they have their
most retarded development among persons, sometimes described as
white-collar workers, who embrace unionism late and slowly. This
appears to be universally true. An examination of the extent of
labor organization among these two types of employees in 1910
and 1920 shows this to have been the case in the United States as
well. While the percentage of total employees who are members
of unions is, roughly, 20, all of the important manual labor groups
stood far above this level; and in 1910 when the general percentage of organization was approximately 10, the same groups of
TABLE 14. — PER CENT OF WAGE EARNERS ORGANIZED IN MAJOR
DIVISIONS OF INDUSTRY
1920 AND 1910
PER CENT ORGANIZED
DIvIsIoN OF
Extraction of Minerals
Manufacturing Industries
Transportation
Building Trades
Stationary Engineers
Stationary Firemen
Trade
Professional Service
Clerical Occupations
Domestic and Personal Service
Public Service
1920
1910
41.0
23.2
37.3
25.5
12.4
19.9
27.3
11.6
1.1
1.0
5.4
4.6
8.3
3.8
7.3
2.0
2.5
17.1
16.4
4.6
9.6
1.8
manual laborers all showed a higher percentage of organization.
The total percentage of organization is in both census years considerably reduced by the absence of many large labor organizations in trade, professional service, clerical occupations, domestic
and personal service, and public service. Membership in the manual workers' groups represented almost 28 and 15 per cent of the
THE EXTENT OF LABOR ORGANIZATION
87
wage earners in those groups in 1920 and 1910; whereas member-
ship among the non-factory workers was in the same years less
than 5 and 2 per cent of all persons engaged in those occupations.
Growth, from 1910 to 1920, occurred in all groups, but it was
greatest in manufacturing, transportation, clerical occupations, and
public service; although the rise in mining and in the building
trades was also considerable.
Differences in the extent of organization among the industries
that comprise these major divisions are quite as striking and as
significant as the differences among the maj or divisions themselves.
At the same time that the whole mining group showed an increase
in the extent of organization from 27 to 41, coal mines increased
their organization from 35 per cent to nearly 51, while trade unionism in copper, gold and other mines actually had a lower percentage of organization in 1920 than in 1910. The strength of the
Western Federation of Miners has not been regained by its successor, the Mine, Mill and Smelter Workers' Union. In 1910 the
workers in the salt, oil and natural gas industry had no union at
all; in 1920 there was a substantial organization with a membership of over 20,000. Such analysis can be pushed even further.
Thus the average number of coal miners in the United States in
1920 was 784,621, of which 639,547 were bituminous and 145,074
anthracite miners.1 It is known that the anthracite miners have a
much higher degree of organization than the soft coal miners.
Bituminous miners were, therefore, in 1920 probably less than 50
per cent organized.
Extent of organization in manufacturing industries runs the whole
gamut from less than 1 per cent of organization in the chemical
and allied industries to more than 57 per cent in clothing. The
tremendous rise in the percentage of organization in this industry
is, in fact, the most striking phenomenon in the whole group of
manufacturing industries. The clothing industry was converted
from one of the weakly organized industries in 1910 into one of
the most strongly organized in 1920. This is attributable, mainly,
as was shown in the discussion of the growth of membership in
this group, to the rise of the International Ladies' Garment Workers
1 U. S. Geological Survey, Mineral Resources of the United States, Part H. "Coal in
1919, 1920, and 1921," p. 494.
88
THE GROWTH OF TRADE UNIONS, 1880-1923
TABLE 15. — PER CENT OF ORGANIZATION AMONG DIVISIONS OF
MANUFACTURING INDUSTRIES
1920 AND 1910
PER CENT ORGANIZED
INDUSTRY
Manufacturing Industries
ChemicalandAllied
Clay, Glass and Stone
Clothing
Food and Kindred Products
Iron and Steel
1920
1910
23.2
11.8
0.2
21.5
57.8
19.4
1.4
20.5
18.9
7.6
28.1
29.4
LiquorandBeverage
....
10.4
14.6
Lumber
18.1
67.6
10.3
Metal (except Iron and Steel)
12.9
6.5
Paper and F'u.lp
7.9
50.1
2.6
34.3
15.0
29.2
3.7
26.9
Printing and Publishing
Textile
Cigar and Tobacco
after their strike in 1910 and to the rapid increase in membership
of the Amalgamated Clothing Workers, after their organization in
the last months of 1914.
The group of clay, glass and stone industries, although it experienced apparently a very slight gain in this decade really suffered
a substantial loss in one of its constituents. Extent of organization in the glass industry fell from 34.2 per cent in 1910 to 27.9
in 1920; at the same time marble and stone yards increased only
very slightly from 45.4 to 47.7 per cent, while potteries and brick,
tile and terra cotta factories made more substantial gains. The
latter were, however, hardly organized in 1910 and even in 1920
had achieved organization of only 9 per cent.
The very large rise in the food group was due almost wholly
to an enormous growth in the extent of organization in slaughter
and packing houses, or, in other words, in the packing industry.
This industry had organization in 1910 of something over 6 per
cent; but in 1920 it had grown to nearly 60 per cent. Butter and
THE EXTENT OF LABOR ORGANIZATION
89
cheese factories, candy factories, flour and grain mills, sugar refin-
èries had practically no organization at all, either in 1910 or in
1920.
Bakeries were .less than one-fifth organized.
Figures for the iron and steel industry appear to contradict current conceptions of the status of labor organization in that industry and are somewhat misleading, although the growth from 1910
to 1920 is a real one. In that portion of the steel industry which
manufactures basic iron and steel products, there was very little
labor organization in either 1910 or 1920, in spite of the fact that
the membership of the Amalgamated Association of Iron, Steel and
Tin Workers, the most important union in this branch of the industry, was much larger in 1920 than in 1910. The category of
iron and steel industries here presented, however, includes the
manufacture of iron and steel products, such as agricultural implements, automobiles, railroad cars, ships and boats, as well as individual occupations such as blacksmiths, boilermakers, pattern
makers, iron molders, machinists. In these branches of the industry there was substantial organization in both census years,
and marked growth from one to the other; but it is unfortunate
that the form of the trade union statistics does not permit the calculation of the percentage organized in each of these branches of
the iron and steel industry. The figures for metal, except iron
and steel, suffer from much, the same defect of representing too
conglomerate an industry.
The classification of the liquor and beverage industries followed
by the census in 1910 and 1920 seems to be so radically different,
that there appears to be no sound basis of comparison. In 1910
this was one of the most strongly organized industries in the country.
While the brewery workers' union is still in existence, it is now a
weak organization, known as the United Brewery, Flour, Cereal
and Soft Drink Workers' Union, with little more than 15,000 members in the soft drink industry, 1,000 in the cider, vinegar, yeast
and alcohol industries, 800 in the flour and cereal industries, and
150 in the syrup industry.
With few exceptions those parts of the manufacturing industries
which were well organized in 1910 had strengthened their organization by 1920. The printing and publishing industry which was a
little better than one-third organized in 1910 achieved organization
90
THE GROWTH OF TRADE UNIONS, 1880-1923
of more than one-half in 1920. The great basic industries like
chemicals, iron and steel, food products, lumber, metals, paper and
pulp, and textiles were, after the total growth from 1910 to 1920,
even in the most favorable instances less than one-fifth organized.
The striking exceptions were, on the one hand, the clothing and
packing industries, in which organization attained an entirely new
and higher level, and, on the other, the liquor and beverage industries where unëommon, but well-known circumstances, prevailed,
that led to disintegration of the union.
Transportation industries show in general a higher level of organization than the manufacturing industries. In fact, the three most
substantial elements of this group were all more than 50 per cent
organized in 1920 and had more than doubled their percentage of
organization since 1910. The most striking change took place in
the division of water transportation where the percentage organized rose from less than 30 to more than 85, due very largely to
spectacular leaps in membership among longshoremen and seamen.
TABLE 16. — PER CENT OF ORGANIZATION AMONG DIVISIONS OF THE
TRANSPORTATION INDUSTRY
1920 AND 1910
•
CENT ORGANIZED
INDUSTRY
1920
1910
37.3
17.1
Water Transportation
Steam Railroads
Electric and Street Railways
85.5
57.5
52.9
28.9
23.5
21.8
Telegraph and Telephone
Post
Teamsters and Chauffeurs
Construction of Streets
25.4
24.8
10.2
31.6
7.0
2.4
All Transportation
11.9
8.3
The very substantial growth of membership among telegraph and
telephone workers came from the fact that the railroad telegraphers'
membership was in 1920 three times that of 1910; whereas the
membership of the commercial telegraphers' union was insignifi-
cant in both years. Another important factor in this situation
THE EXTENT OF LABOR ORGANIZATION
91
was the establishment, a few years after 1910, of a union among
women telephone operators. The figures for the extent of organization among teamsters and chauffeurs are probably underestimates in both census years because of the great practical difficulties
involved in separating this group into the employers and wage
earners and in calculating the percentage organized of wage earners
alone.
Labor organization in professional service is restricted almost
completely to the theatre and to musicians, although there has
recently grown up a small union of engineering draftsmen. Among
the clerical workers the principal source of growth since 1910 is
the rapid spread of organization among the railway clerks. Unionism in domestic and personal service is in 1920 as in 1910 limited
to fair, organization among barbers and waiters and to exceedingly
slim organization among laundry workers.
All of these figures naturally raise interesting questions as to
their significance in estimating the relative strength of labor organization in various occupations and industries. High and low
percentages of organization are not necessarily synonymous with
strength and weakness, and need, in fact, to be interpreted with
some reference to the nature of the industry in which. the unions
operate and to the constitution of the particular union. In general, percentages of organization by industry are misleading unless
the fact that the large bulk of American unions are occupational
or trade unions is taken into consideration. Because of this highly
significant characteristic, union membership in any industry does
not represent the membership of a single union claiming jurisdic-
tion over all the employees in that industry, but it is really an
aggregate of the memberships of many unions, some limiting their
jurisdiction to the workers in a single skilled craft and others admitting a more diversified lot of semi-skilled and unskilled workers.
The only outstanding exceptions to this rule are the mining and
clothing unions, which are essentially industrial unions.
The unions of skilled craft workers are, also, the older and stronger
organizations and their membership is on the whole less subject to
fluctuation. The unions of semi-skilled and unskilled workers are
more recently organized, weaker and more sensitive to the strains
92
THE GROWTH OF TRADE UNIONS, 1880-1923
imposed by industrial depression. When, accordingly, these diverse
classes of workers are combined into one industrial category, the real
strength of the skilled unions is, to a degree, concealed in the gross
results. This is particularly true of the building trades and steam
railroad industry, where the backbone of unionism has for a long
time been the relatively few organizations of skilled craftsmen. The
following table shows clearly the divergence between the extent of
TABLE 17. — PER CENT OF ORGANIZATION AMONG SELECTED OCCU-
PATIONS IN THE BUILDING TRADES
1920 AND 1910
PER CEN'r
OCCUPATION
Brick and Stone Masons
Carpenters and Joiners
Painters, etc
Plasterers
Plumbers and Gas Fitters
1920
1910
40.5
29.1
46.6
33.5
39.1
20.8
17.6
32.0
20.7
organization for the building trades as a whole and among a few
skilled crafts. Thus in 1920, when all employees in the building
trades were just about one-fourth organized, the bricklayers, carpenters and plasterers were about 50 per cent, and the plumbers
more than one-third organized. The same disparities existed in
1910. At that time the whole industry was one-sixth organized,
but the bricklayers and plasterers were about one-third organized.
On the steam railroads, likewise, even the very high percentage
of organization for the industry as a whole, 56.7 per cent, was
exceeded by the percentages of organization for railway conductors,
locomotive engineers, and locomotive firemen.1 Census statistics
for the groups of locomotive firemen and enginemen are unfortunately not reliable, because of the confusion in enumeration between
stationary and locomotive firemen and engineers. This confusion
led to an overestimate in the number of locomotive engineers as
compared with the locomotive firemen and probably a gross over1
Tables
VIII and IX.
THE EXTENT OF LABOR• ORGANIZATION
93
estimate of both groups.' Union membership statistics are likewise defective, since the Brotherhood of Locomotive Firemen, which
because of its insurance features retains as members firemen who
have become engineers and who are, consequently, also members
of the engineers' union, was unable to separate its membership into
engineers and firemen. The figures as they now stand contain
some double counting in the membership of the firemen's union.
It is known, however, that both the engineers and firemen were well
over 75 per cent organized in 1920. Railway conductors are likewise nearly 100 per cent organized, although the statistics indicate
an organization of only 72 per cent, less than in 1910. The discrepancy between 1910 and 1920 is due in large part to the failure
of the Brotherhood of Railroad Trainmen to report the number
of its members who were railway conductors. In 1910 this number amounted to 13,000; union membership for 1920 is, consequeritly, understated. Furthermore, the census reports a larger
number of railway conductors than does the Interstate Commerce
Commission.2 According to this agency the average number of
conductors on class I railroads in the year ending December 31,
1920 was 58,321, whereas the census figure for early January, 1920
is 74,539. It is possible that a portion of this difference may be
due to the reporting by the census of electric railway conductors
as working on steam
Another factor that should be taken into account in judging the
strength of the labor movement in the United States is the size
and infinite variety of the country. The enormous expense of conducting organization campaigns and of maintaining a staff of organizers all over the country, as well as the inherent difficulty of
organizing a thin and scattered industrial population has concen-
trated trade unions, in many industries, in the large industrial
cities. There is, moreover, considerable evidence, although the
supporting data are not available, that labor organization is much
more widespread in certain sections of the country than in others.
The Census of Occupations reports for 1920, 91,345 locomotive firemen and 109,899
locomotive engineers. The Interstate Commerce Commission, however (Annual Re1
port on Stattsttcs of Railways in the United States, 1920, pp. ix, xix, xx), reports the
average number of locomotive engineers, on class I railroads, in the year ending December 31, 1920, to be 67,887 and the average number of locomotive firemen, 69,935.
2
Annual Report on Statistics of Railways in the United States, 1920, p. 20.
Census of Occupations, 1920, p. 16.
THE GROWTH OF TRADE UNIONS, 1880-1923
94
The South as a whole, for example, even in its industrial centers
is very thinly organized, if at all; whereas the industrial East
would probably show a high percentage of organization. The almost complete absence of trade unions in the textile industry in
the South brings down the percentage of organization for the textile industry as a whole, although some of its centers in New England are tolerably well organized.
Equally interesting• comparisons
could be made for other industries and for other sections of the
country. But the unions are either unwilling or unable to submit
a detailed geographical distribution of their membership.
Two interesting samples, which throw some light on these phenomena, have been collected for the occupations of bricklaying
and printing. In the printing trade the largest source of union
membership is the newspaper office: whereas the union is weaker
in the book and jobbing trade and probably has a very light membership among compositors in small towns, where there are a considerable number of one-man shops. In 1920, compositors, linotypers and typesetters in the United States were 46.4 per cent
organized. The next table shows for 1920 the number of compositors in a list of selected cities, the membership of the Typographical Union in those cities and the percentage of organization.
Except where membership was affected by peculiar circumstances,
TABLE 18. — PER CENT OF ORGANIZATION AMONG COMPOSITORS,
LINOTYPERS AND TYPESETTERS IN SELECTED CITIES
1920
NUMBER OF CoMPosIToRs,
LINOTYPERS AND
CITY
TYPESETTERS
New York
Chicago
..
Philadelphia....
St. Louis
San
Francisco. .
Baltimore
Cleveland
Boston.
a Census
.
. ...
.
..
MEMBERSEIP
o'
PER CENT
ORGANIZED
21,429
10,907
9,044
5,119
42.2
46.9
5,708
1,606
28.1
2,205
1,323
60.0
1,457
1,257
86.2
1,886
1,741
2,713
898
47.6
59.3
77.3
1,033
2,098
of Occupations, 1920.
'From central office of the International Typographical Union.
THE EXTENT OF LABOR ORGANIZATION
95
the extent of organization was higher than in the whole country.
Thus in St. Louis, San Francisco, Cleveland and Boston it was
considerably higher. Philadelphia, which has the lowest per cent
of organization, is notoriously low in the scale of organization in• all
industries. Percentages are lowered in New York and Chicago by
lack of control over book and job printing establishments in particular and over the small printing shops in general; while in
Baltimore the union has not recovered from the weakening effect
of the strike for the 44-hour week of a few years ago, in which it
lost many members.
Similar data for the bricklaying trade, contained in this next
tabulation, is even more convincing on this point. In every large
city the percentage of organization was considerably greater than
the 50 per cent for the entire United States. Obvious discrepancies
TABLE 19. — PER CENT OF ORGANIZATION AMONG BRICK AND STONE
MASONS IN SELECTED CITIES
1920
CITY
Chicago
Baltimore
Boston
Cleveland
.
New York
Philadelphia
Pittsburgh
a Census
NUMBER OF
MEMBERSHIP
BRICK AND STONE
MASONSO
OP
PER CENT
ORGANIZED
5,303
1,194
1,274
2,351
4,229
927
1,220
1,866
79.7
77.6
95.8
79.4
9,985
3,818
1,159
5,925
2,188
1,273
59.3
57.3
....
of Occupations, 1920.
First Biennial and
53d Report of the President, Secretary, an.,d Official Auditor, Bricklayers, Masons and
Plasterers' Union, 1920.
b
Average monthly membership for year ended June 30, 1920.
in the table, such as the excess in the membership of the union in
Pittsburgh over the number of masons in the city, may be due to
the fact that the census figures are as of early January while the
union statistics are the average for the fiscal year. The appreciably lower percentage of organization in New York City may
be explained by the lack of union control over the many small
building operations in the outlying areas of the city.
96
THE. GROWTH OF TRADE UNIONS, 1880-1923
More than three years have elapsed since the taking of the census of occupations of 1920. It is known that in this period trade
unions lost heavily in membership. The trend in the size of the
working population during the same period is still a matter of
speculation.
Aggregate statistics of the 1923 census of manufac-
tures are not available at this writing. What evidence there is
would seem to indicate a smaller number of employees in manufacturing industries in 1923 than in 1920. The statistics of the
Interstate Commerce Commission show fewer persons employed on
steam railroads during 1923 than during 1920. How large was the
total decline, if any, in the number of employees of the country
in these last years cannot be estimated. All things considered,
however, it is probable that the extent of organization for all industry was considerably greater in 1923 than in the years immediately before and after the declaration of the World War;, that
for manufacturing industries it is substantially less in 1923 than
in 1920; that in transportation and mining the drop from 1920
to 1923 is not so great as in manufacturing industries; and that
in the building trades, the drop in these last years was slight and
organization in that industry in 1923 stood little, if at all, below
1920.
Any forecast of the trend of union organization in the future
must reckon with two conditions that are comparatively strange
in the American industrial situation. The first of these is legislation restricting immigration into the United States. The immigration law of 1924 establishes immigrant quotas which may reduce
enormously the flow of immigrant labor into the country. The
effect of this restrictive measure has already been noticed as one
of the causes of the growth of labor organization in the clothing
industry. It may be expected to exert the same kind of influence
in other industries as well. The second factor is not so tangible
and has to do with the probable influence in the future of the impetus given the movement by large gains made since 1915.
Already,
even in the cases where heavy losses have been registered since
1920, there is some evidence of the consolidation by labor organizations of at least a portion of their advances. How potent a force
this impetus is, it is hazardous to guess; but it is easy to underestimate the influence of intangible social forces of this kind.